TY - CONF
AB - In the standard framework for worst-case execution time (WCET) analysis of programs, the main data structure is a single instance of integer linear programming (ILP) that represents the whole program. The instance of this NP-hard problem must be solved to find an estimate forWCET, and it must be refined if the estimate is not tight.We propose a new framework for WCET analysis, based on abstract segment trees (ASTs) as the main data structure. The ASTs have two advantages. First, they allow computing WCET by solving a number of independent small ILP instances. Second, ASTs store more expressive constraints, thus enabling a more efficient and precise refinement procedure. In order to realize our framework algorithmically, we develop an algorithm for WCET estimation on ASTs, and we develop an interpolation-based counterexample-guided refinement scheme for ASTs. Furthermore, we extend our framework to obtain parametric estimates of WCET. We experimentally evaluate our approach on a set of examples from WCET benchmark suites and linear-algebra packages. We show that our analysis, with comparable effort, provides WCET estimates that in many cases significantly improve those computed by existing tools.
AU - Cerny, Pavol
AU - Henzinger, Thomas A
AU - Kovács, Laura
AU - Radhakrishna, Arjun
AU - Zwirchmayr, Jakob
ID - 1836
TI - Segment abstraction for worst-case execution time analysis
VL - 9032
ER -
TY - JOUR
AB - Transition to turbulence in straight pipes occurs in spite of the linear stability of the laminar Hagen-Poiseuille flow if both the amplitude of flow perturbations and the Reynolds number Re exceed a minimum threshold (subcritical transition). As the pipe curvature increases, centrifugal effects become important, modifying the basic flow as well as the most unstable linear modes. If the curvature (tube-to-coiling diameter d/D) is sufficiently large, a Hopf bifurcation (supercritical instability) is encountered before turbulence can be excited (subcritical instability). We trace the instability thresholds in the Re - d/D parameter space in the range 0.01 ≤ d/D\ ≤ 0.1 by means of laser-Doppler velocimetry and determine the point where the subcritical and supercritical instabilities meet. Two different experimental set-ups are used: a closed system where the pipe forms an axisymmetric torus and an open system employing a helical pipe. Implications for the measurement of friction factors in curved pipes are discussed.
AU - Kühnen, Jakob
AU - Braunshier, P
AU - Schwegel, M
AU - Kuhlmann, Hendrik
AU - Hof, Björn
ID - 1837
IS - 5
JF - Journal of Fluid Mechanics
TI - Subcritical versus supercritical transition to turbulence in curved pipes
VL - 770
ER -
TY - CONF
AB - Synthesis of program parts is particularly useful for concurrent systems. However, most approaches do not support common design tasks, like modifying a single process without having to re-synthesize or verify the whole system. Assume-guarantee synthesis (AGS) provides robustness against modifications of system parts, but thus far has been limited to the perfect information setting. This means that local variables cannot be hidden from other processes, which renders synthesis results cumbersome or even impossible to realize.We resolve this shortcoming by defining AGS under partial information. We analyze the complexity and decidability in different settings, showing that the problem has a high worstcase complexity and is undecidable in many interesting cases. Based on these observations, we present a pragmatic algorithm based on bounded synthesis, and demonstrate its practical applicability on several examples.
AU - Bloem, Roderick
AU - Chatterjee, Krishnendu
AU - Jacobs, Swen
AU - Könighofer, Robert
ID - 1838
TI - Assume-guarantee synthesis for concurrent reactive programs with partial information
VL - 9035
ER -
TY - CONF
AB - We present MultiGain, a tool to synthesize strategies for Markov decision processes (MDPs) with multiple mean-payoff objectives. Our models are described in PRISM, and our tool uses the existing interface and simulator of PRISM. Our tool extends PRISM by adding novel algorithms for multiple mean-payoff objectives, and also provides features such as (i) generating strategies and exploring them for simulation, and checking them with respect to other properties; and (ii) generating an approximate Pareto curve for two mean-payoff objectives. In addition, we present a new practical algorithm for the analysis of MDPs with multiple mean-payoff objectives under memoryless strategies.
AU - Brázdil, Tomáš
AU - Chatterjee, Krishnendu
AU - Forejt, Vojtěch
AU - Kučera, Antonín
ID - 1839
TI - Multigain: A controller synthesis tool for MDPs with multiple mean-payoff objectives
VL - 9035
ER -
TY - JOUR
AB - In this paper, we present a method for reducing a regular, discrete-time Markov chain (DTMC) to another DTMC with a given, typically much smaller number of states. The cost of reduction is defined as the Kullback-Leibler divergence rate between a projection of the original process through a partition function and a DTMC on the correspondingly partitioned state space. Finding the reduced model with minimal cost is computationally expensive, as it requires an exhaustive search among all state space partitions, and an exact evaluation of the reduction cost for each candidate partition. Our approach deals with the latter problem by minimizing an upper bound on the reduction cost instead of minimizing the exact cost. The proposed upper bound is easy to compute and it is tight if the original chain is lumpable with respect to the partition. Then, we express the problem in the form of information bottleneck optimization, and propose using the agglomerative information bottleneck algorithm for searching a suboptimal partition greedily, rather than exhaustively. The theory is illustrated with examples and one application scenario in the context of modeling bio-molecular interactions.
AU - Geiger, Bernhard
AU - Petrov, Tatjana
AU - Kubin, Gernot
AU - Koeppl, Heinz
ID - 1840
IS - 4
JF - IEEE Transactions on Automatic Control
SN - 0018-9286
TI - Optimal Kullback-Leibler aggregation via information bottleneck
VL - 60
ER -
TY - JOUR
AB - We propose a new family of message passing techniques for MAP estimation in graphical models which we call Sequential Reweighted Message Passing (SRMP). Special cases include well-known techniques such as Min-Sum Diffusion (MSD) and a faster Sequential Tree-Reweighted Message Passing (TRW-S). Importantly, our derivation is simpler than the original derivation of TRW-S, and does not involve a decomposition into trees. This allows easy generalizations. The new family of algorithms can be viewed as a generalization of TRW-S from pairwise to higher-order graphical models. We test SRMP on several real-world problems with promising results.
AU - Kolmogorov, Vladimir
ID - 1841
IS - 5
JF - IEEE Transactions on Pattern Analysis and Machine Intelligence
TI - A new look at reweighted message passing
VL - 37
ER -
TY - JOUR
AB - Based on extrapolation from excitatory synapses, it is often assumed that depletion of the releasable pool of synaptic vesicles is the main factor underlying depression at inhibitory synapses. In this issue of Neuron, using subcellular patch-clamp recording from inhibitory presynaptic terminals, Kawaguchi and Sakaba (2015) show that at Purkinje cell-deep cerebellar nuclei neuron synapses, changes in presynaptic action potential waveform substantially contribute to synaptic depression. Based on extrapolation from excitatory synapses, it is often assumed that depletion of the releasable pool of synaptic vesicles is the main factor underlying depression at inhibitory synapses. In this issue of Neuron, using subcellular patch-clamp recording from inhibitory presynaptic terminals, Kawaguchi and Sakaba (2015) show that at Purkinje cell-deep cerebellar nuclei neuron synapses, changes in presynaptic action potential waveform substantially contribute to synaptic depression.
AU - Vandael, David H
AU - Espinoza Martinez, Claudia M
AU - Jonas, Peter M
ID - 1845
IS - 6
JF - Neuron
TI - Excitement about inhibitory presynaptic terminals
VL - 85
ER -
TY - JOUR
AB - Modal transition systems (MTS) is a well-studied specification formalism of reactive systems supporting a step-wise refinement methodology. Despite its many advantages, the formalism as well as its currently known extensions are incapable of expressing some practically needed aspects in the refinement process like exclusive, conditional and persistent choices. We introduce a new model called parametric modal transition systems (PMTS) together with a general modal refinement notion that overcomes many of the limitations. We investigate the computational complexity of modal and thorough refinement checking on PMTS and its subclasses and provide a direct encoding of the modal refinement problem into quantified Boolean formulae, allowing us to employ state-of-the-art QBF solvers for modal refinement checking. The experiments we report on show that the feasibility of refinement checking is more influenced by the degree of nondeterminism rather than by the syntactic restrictions on the types of formulae allowed in the description of the PMTS.
AU - Beneš, Nikola
AU - Kretinsky, Jan
AU - Larsen, Kim
AU - Möller, Mikael
AU - Sickert, Salomon
AU - Srba, Jiří
ID - 1846
IS - 2-3
JF - Acta Informatica
TI - Refinement checking on parametric modal transition systems
VL - 52
ER -
TY - JOUR
AU - Grones, Peter
AU - Friml, Jiřĺ
ID - 1847
IS - 3
JF - Molecular Plant
TI - ABP1: Finally docking
VL - 8
ER -
TY - JOUR
AB - The ability to escape apoptosis is a hallmark of cancer-initiating cells and a key factor of resistance to oncolytic therapy. Here, we identify FAM96A as a ubiquitous, evolutionarily conserved apoptosome-activating protein and investigate its potential pro-apoptotic tumor suppressor function in gastrointestinal stromal tumors (GISTs). Interaction between FAM96A and apoptotic peptidase activating factor 1 (APAF1) was identified in yeast two-hybrid screen and further studied by deletion mutants, glutathione-S-transferase pull-down, co-immunoprecipitation and immunofluorescence. Effects of FAM96A overexpression and knock-down on apoptosis sensitivity were examined in cancer cells and zebrafish embryos. Expression of FAM96A in GISTs and histogenetically related cells including interstitial cells of Cajal (ICCs), “fibroblast-like cells” (FLCs) and ICC stem cells (ICC-SCs) was investigated by Northern blotting, reverse transcription—polymerase chain reaction, immunohistochemistry and Western immunoblotting. Tumorigenicity of GIST cells and transformed murine ICC-SCs stably transduced to re-express FAM96A was studied by xeno- and allografting into immunocompromised mice. FAM96A was found to bind APAF1 and to enhance the induction of mitochondrial apoptosis. FAM96A protein or mRNA was dramatically reduced or lost in 106 of 108 GIST samples representing three independent patient cohorts. Whereas ICCs, ICC-SCs and FLCs, the presumed normal counterparts of GIST, were found to robustly express FAM96A protein and mRNA, FAM96A expression was much reduced in tumorigenic ICC-SCs. Re-expression of FAM96A in GIST cells and transformed ICC-SCs increased apoptosis sensitivity and diminished tumorigenicity. Our data suggest FAM96A is a novel pro-apoptotic tumor suppressor that is lost during GIST tumorigenesis.
AU - Schwamb, Bettina
AU - Pick, Robert
AU - Fernández, Sara
AU - Völp, Kirsten
AU - Heering, Jan
AU - Dötsch, Volker
AU - Bösser, Susanne
AU - Jung, Jennifer
AU - Beinoravičiute Kellner, Rasa
AU - Wesely, Josephine
AU - Zörnig, Inka
AU - Hammerschmidt, Matthias
AU - Nowak, Matthias
AU - Penzel, Roland
AU - Zatloukal, Kurt
AU - Joos, Stefan
AU - Rieker, Ralf
AU - Agaimy, Abbas
AU - Söder, Stephan
AU - Reid Lombardo, Kmarie
AU - Kendrick, Michael
AU - Bardsley, Michael
AU - Hayashi, Yujiro
AU - Asuzu, David
AU - Syed, Sabriya
AU - Ördög, Tamás
AU - Zörnig, Martin
ID - 1848
IS - 6
JF - International Journal of Cancer
TI - FAM96A is a novel pro-apoptotic tumor suppressor in gastrointestinal stromal tumors
VL - 137
ER -
TY - JOUR
AB - Cell polarity is a fundamental property of pro- and eukaryotic cells. It is necessary for coordination of cell division, cell morphogenesis and signaling processes. How polarity is generated and maintained is a complex issue governed by interconnected feed-back regulations between small GTPase signaling and membrane tension-based signaling that controls membrane trafficking, and cytoskeleton organization and dynamics. Here, we will review the potential role for calcium as a crucial signal that connects and coordinates the respective processes during polarization processes in plants. This article is part of a Special Issue entitled: 13th European Symposium on Calcium.
AU - Himschoot, Ellie
AU - Beeckman, Tom
AU - Friml, Jiřĺ
AU - Vanneste, Steffen
ID - 1849
IS - 9
JF - Biochimica et Biophysica Acta - Molecular Cell Research
TI - Calcium is an organizer of cell polarity in plants
VL - 1853
ER -
TY - JOUR
AB - Entomopathogenic fungi are potent biocontrol agents that are widely used against insect pests, many of which are social insects. Nevertheless, theoretical investigations of their particular life history are scarce. We develop a model that takes into account the main distinguishing features between traditionally studied diseases and obligate killing pathogens, like the (biocontrol-relevant) insect-pathogenic fungi Metarhizium and Beauveria. First, obligate killing entomopathogenic fungi produce new infectious particles (conidiospores) only after host death and not yet on the living host. Second, the killing rates of entomopathogenic fungi depend strongly on the initial exposure dosage, thus we explicitly consider the pathogen load of individual hosts. Further, we make the model applicable not only to solitary host species, but also to group living species by incorporating social interactions between hosts, like the collective disease defences of insect societies. Our results identify the optimal killing rate for the pathogen that minimises its invasion threshold. Furthermore, we find that the rate of contact between hosts has an ambivalent effect: dense interaction networks between individuals are considered to facilitate disease outbreaks because of increased pathogen transmission. In social insects, this is compensated by their collective disease defences, i.e., social immunity. For the type of pathogens considered here, we show that even without social immunity, high contact rates between live individuals dilute the pathogen in the host colony and hence can reduce individual pathogen loads below disease-causing levels.
AU - Novak, Sebastian
AU - Cremer, Sylvia
ID - 1850
IS - 5
JF - Journal of Theoretical Biology
TI - Fungal disease dynamics in insect societies: Optimal killing rates and the ambivalent effect of high social interaction rates
VL - 372
ER -
TY - JOUR
AB - We consider mating strategies for females who search for males sequentially during a season of limited length. We show that the best strategy rejects a given male type if encountered before a time-threshold but accepts him after. For frequency-independent benefits, we obtain the optimal time-thresholds explicitly for both discrete and continuous distributions of males, and allow for mistakes being made in assessing the correct male type. When the benefits are indirect (genes for the offspring) and the population is under frequency-dependent ecological selection, the benefits depend on the mating strategy of other females as well. This case is particularly relevant to speciation models that seek to explore the stability of reproductive isolation by assortative mating under frequency-dependent ecological selection. We show that the indirect benefits are to be quantified by the reproductive values of couples, and describe how the evolutionarily stable time-thresholds can be found. We conclude with an example based on the Levene model, in which we analyze the evolutionarily stable assortative mating strategies and the strength of reproductive isolation provided by them.
AU - Priklopil, Tadeas
AU - Kisdi, Eva
AU - Gyllenberg, Mats
ID - 1851
IS - 4
JF - Evolution
TI - Evolutionarily stable mating decisions for sequentially searching females and the stability of reproductive isolation by assortative mating
VL - 69
ER -
TY - JOUR
AB - The traditional synthesis question given a specification asks for the automatic construction of a system that satisfies the specification, whereas often there exists a preference order among the different systems that satisfy the given specification. Under a probabilistic assumption about the possible inputs, such a preference order is naturally expressed by a weighted automaton, which assigns to each word a value, such that a system is preferred if it generates a higher expected value. We solve the following optimal synthesis problem: given an omega-regular specification, a Markov chain that describes the distribution of inputs, and a weighted automaton that measures how well a system satisfies the given specification under the input assumption, synthesize a system that optimizes the measured value. For safety specifications and quantitative measures that are defined by mean-payoff automata, the optimal synthesis problem reduces to finding a strategy in a Markov decision process (MDP) that is optimal for a long-run average reward objective, which can be achieved in polynomial time. For general omega-regular specifications along with mean-payoff automata, the solution rests on a new, polynomial-time algorithm for computing optimal strategies in MDPs with mean-payoff parity objectives. Our algorithm constructs optimal strategies that consist of two memoryless strategies and a counter. The counter is in general not bounded. To obtain a finite-state system, we show how to construct an ε-optimal strategy with a bounded counter, for all ε > 0. Furthermore, we show how to decide in polynomial time if it is possible to construct an optimal finite-state system (i.e., a system without a counter) for a given specification. We have implemented our approach and the underlying algorithms in a tool that takes qualitative and quantitative specifications and automatically constructs a system that satisfies the qualitative specification and optimizes the quantitative specification, if such a system exists. We present some experimental results showing optimal systems that were automatically generated in this way.
AU - Chatterjee, Krishnendu
AU - Henzinger, Thomas A
AU - Jobstmann, Barbara
AU - Singh, Rohit
ID - 1856
IS - 1
JF - Journal of the ACM
TI - Measuring and synthesizing systems in probabilistic environments
VL - 62
ER -
TY - CONF
AB - Sharing information between multiple tasks enables algorithms to achieve good generalization performance even from small amounts of training data. However, in a realistic scenario of multi-task learning not all tasks are equally related to each other, hence it could be advantageous to transfer information only between the most related tasks. In this work we propose an approach that processes multiple tasks in a sequence with sharing between subsequent tasks instead of solving all tasks jointly. Subsequently, we address the question of curriculum learning of tasks, i.e. finding the best order of tasks to be learned. Our approach is based on a generalization bound criterion for choosing the task order that optimizes the average expected classification performance over all tasks. Our experimental results show that learning multiple related tasks sequentially can be more effective than learning them jointly, the order in which tasks are being solved affects the overall performance, and that our model is able to automatically discover the favourable order of tasks.
AU - Pentina, Anastasia
AU - Sharmanska, Viktoriia
AU - Lampert, Christoph
ID - 1857
TI - Curriculum learning of multiple tasks
ER -
TY - CONF
AB - We study the problem of predicting the future, though only in the probabilistic sense of estimating a future state of a time-varying probability distribution. This is not only an interesting academic problem, but solving this extrapolation problem also has many practical application, e.g. for training classifiers that have to operate under time-varying conditions. Our main contribution is a method for predicting the next step of the time-varying distribution from a given sequence of sample sets from earlier time steps. For this we rely on two recent machine learning techniques: embedding probability distributions into a reproducing kernel Hilbert space, and learning operators by vector-valued regression. We illustrate the working principles and the practical usefulness of our method by experiments on synthetic and real data. We also highlight an exemplary application: training a classifier in a domain adaptation setting without having access to examples from the test time distribution at training time.
AU - Lampert, Christoph
ID - 1858
TI - Predicting the future behavior of a time-varying probability distribution
ER -
TY - CONF
AB - Structural support vector machines (SSVMs) are amongst the best performing models for structured computer vision tasks, such as semantic image segmentation or human pose estimation. Training SSVMs, however, is computationally costly, because it requires repeated calls to a structured prediction subroutine (called \emph{max-oracle}), which has to solve an optimization problem itself, e.g. a graph cut.
In this work, we introduce a new algorithm for SSVM training that is more efficient than earlier techniques when the max-oracle is computationally expensive, as it is frequently the case in computer vision tasks. The main idea is to (i) combine the recent stochastic Block-Coordinate Frank-Wolfe algorithm with efficient hyperplane caching, and (ii) use an automatic selection rule for deciding whether to call the exact max-oracle or to rely on an approximate one based on the cached hyperplanes.
We show experimentally that this strategy leads to faster convergence to the optimum with respect to the number of requires oracle calls, and that this translates into faster convergence with respect to the total runtime when the max-oracle is slow compared to the other steps of the algorithm.
AU - Shah, Neel
AU - Kolmogorov, Vladimir
AU - Lampert, Christoph
ID - 1859
TI - A multi-plane block-coordinate Frank-Wolfe algorithm for training structural SVMs with a costly max-oracle
ER -
TY - CONF
AB - Classifiers for object categorization are usually evaluated by their accuracy on a set of i.i.d. test examples. This provides us with an estimate of the expected error when applying the classifiers to a single new image. In real application, however, classifiers are rarely only used for a single image and then discarded. Instead, they are applied sequentially to many images, and these are typically not i.i.d. samples from a fixed data distribution, but they carry dependencies and their class distribution varies over time. In this work, we argue that the phenomenon of correlated data at prediction time is not a nuisance, but a blessing in disguise. We describe a probabilistic method for adapting classifiers at prediction time without having to retrain them. We also introduce a framework for creating realistically distributed image sequences, which offers a way to benchmark classifier adaptation methods, such as the one we propose. Experiments on the ILSVRC2010 and ILSVRC2012 datasets show that adapting object classification systems at prediction time can significantly reduce their error rate, even with no additional human feedback.
AU - Royer, Amélie
AU - Lampert, Christoph
ID - 1860
TI - Classifier adaptation at prediction time
ER -
TY - JOUR
AB - Continuous-time Markov chains are commonly used in practice for modeling biochemical reaction networks in which the inherent randomness of themolecular interactions cannot be ignored. This has motivated recent research effort into methods for parameter inference and experiment design for such models. The major difficulty is that such methods usually require one to iteratively solve the chemical master equation that governs the time evolution of the probability distribution of the system. This, however, is rarely possible, and even approximation techniques remain limited to relatively small and simple systems. An alternative explored in this article is to base methods on only some low-order moments of the entire probability distribution. We summarize the theory behind such moment-based methods for parameter inference and experiment design and provide new case studies where we investigate their performance.
AU - Ruess, Jakob
AU - Lygeros, John
ID - 1861
IS - 2
JF - ACM Transactions on Modeling and Computer Simulation
TI - Moment-based methods for parameter inference and experiment design for stochastic biochemical reaction networks
VL - 25
ER -
TY - JOUR
AB - The Altshuler–Shklovskii formulas (Altshuler and Shklovskii, BZh Eksp Teor Fiz 91:200, 1986) predict, for any disordered quantum system in the diffusive regime, a universal power law behaviour for the correlation functions of the mesoscopic eigenvalue density. In this paper and its companion (Erdős and Knowles, The Altshuler–Shklovskii formulas for random band matrices I: the unimodular case, 2013), we prove these formulas for random band matrices. In (Erdős and Knowles, The Altshuler–Shklovskii formulas for random band matrices I: the unimodular case, 2013) we introduced a diagrammatic approach and presented robust estimates on general diagrams under certain simplifying assumptions. In this paper, we remove these assumptions by giving a general estimate of the subleading diagrams. We also give a precise analysis of the leading diagrams which give rise to the Altschuler–Shklovskii power laws. Moreover, we introduce a family of general random band matrices which interpolates between real symmetric (β = 1) and complex Hermitian (β = 2) models, and track the transition for the mesoscopic density–density correlation. Finally, we address the higher-order correlation functions by proving that they behave asymptotically according to a Gaussian process whose covariance is given by the Altshuler–Shklovskii formulas.
AU - Erdös, László
AU - Knowles, Antti
ID - 1864
IS - 3
JF - Annales Henri Poincare
TI - The Altshuler–Shklovskii formulas for random band matrices II: The general case
VL - 16
ER -